Microsoft = Roach Motel?

The August 18-25 double issue of Newsweek, whose theme is “The Future of Technology”, contains a brief interview with Paul Saffo, identified only as being from the Institute For the Future. Asked “are there any obstacles to innovation?” Saffo responds “We’ve got a couple of gorillas holding back innovation. Microsoft is a big intellectual roach motel. All the big minds go in and they don’t come out.”

Of course Saffo doesn’t address the question of what alternatives the big minds have. If they want to work at a software products company and have a reasonable chance of getting their creation into the hands of customers, the choices pretty much boil down to Microsoft or some company that is likely to be put out of business soon by Microsoft.

[Of course there are quite a few programmers, as distinct from engineers, who are happy with a warm cubicle and a fat salary even if they have no impact on users or their employer’s profits. Sadly, however, in an age where spectacular managerial incompetence continues to be the norm it seems that many managers have gotten smart enough to eliminate tech jobs that aren’t directly profitable.]

15 thoughts on “Microsoft = Roach Motel?

  1. Philip, are you envious that you don’t work with Microsoft Research? They have some VERY big names in computer science: Jim Gray, Rick Rashid, C.A.R. Hoare, Leslie Lamport, Butler Lampson, etc.

  2. If they want their work to be seen and used they should work with open source. If they want their work to be abused by marketing departments they should work with Microsoft.

  3. Marketing abuse? Marketing is making sure their work is sold. I am yet to find anyone that works on the SQL Server, Windows Server and Visual Studio teams complaining about how MS Marketing is making their lives harder.

    I am a programmer and like every professional athlete I would rather play with the team with the winning streak and the big bucks. I mean, this is a no-brainer. Would I rather work for a small company that is struggling to make sales and because of this makes all sorts of horrendous compromises just to make a sale? Or would I prefer to work for a giant company that is willing to pay me a hell of a lot of money and put incredible resources at my disposal to do the same kind of job, and this without having to feel the pressure that the company is not making any money and layoffs are imminent.

    But yes, doing open source work is a very quick way to satisfy the ego. Write something cool, give it away and you will still be poor but will on the other hand be worshipped accordingly. I don’t care about recognition, I just want to be able to write software without having to worry about crap like the customer wanting icons in cornsilk blue more than having the application be bug-free.

  4. Scott: I think you’ve restated Saffo’s point, i.e., lots of top computer nerds couldn’t find any better place to work than MSFT. And I’ve made a few visits to Microsoft Research to give talks or see friends, always being impressed by something that they’ve done there. Am I therefore envious of people who work at Microsoft Research? No, but mostly because I’m not envious of anyone who has to work at all! I prefer to be here in Bryce Canyon, relaxing after a helicopter tour (below the rim with all the gusto of a turbine engine; the Sierra Club has managed to rid the trails of Golden Retrievers but not of jet-powered aircraft 300′ overhead), rather than at a desk in Redmond.

    If I were looking for a job, would I want to work at Microsoft Research? No. First, Microsoft offices are dog-unfriendly (they have receptionists/thugs at the front doors to keep out dogs and other undesirables). Second, the best research universities are more fun. It is conceivable that Microsoft will eventually assemble a better collection of computer scientists than any university. But so what? Do you want to go to a cocktail party with a bunch of computer scientists?

    The MIT community is full of fascinating intelligent people in exciting fields such as biology, physics, aeronautical engineering, geology/meteorology, mechanical engineering, chemistry, materials science, etc. MIT is a place where you can get an informed answer to any technical or scientific question, usually from someone with a much greater spirit of fun and inquiry than the average computer nerd. And MIT is right next to Harvard so you can play poker with the Law School faculty and use their blog server. Finally universities are blessed with a lot of young people. Corporate slaves earn high-ish salaries and are useful for shoveling profits into the hands of top managers and, occasionally, shareholders, but they are generally old, boring, and on the downslope of the IQ curve. Undergrads and graduate students are young, refreshing, and on the upslope of intelligence and learning. At lunch, they talk about what they learned that morning or what entertainment they enjoyed the night before rather than about where they are taking their kids for the weekend or how their 401k plan is doing.

  5. Philip, I have both worked at Microsoft (as a Software Engineer, not in the Research department) and, more recently, went back to school to get a Masters in Computer Science, so I feel I have seen both worlds “from the inside.” I also regularly talk with both current MS employees and friends who are either working towards their PhDs or are professors.

    You bring up some good points about the benefits of working in a university, and do point out some of the disadvantages of working for MS. But you fail to point out the disadvantages of working in academia – beauracracy, meetings out the wazoo, politics, etc. – and leave out the benefits of working at MS – flex hours, job security, abundant pay, working with people who are passionaite about what they do, etc.

    Your statements that universities tend to have a buzz of excitement around them from undergrads/grads/and professions excited at learning, I agree this is true in most university settings. I also found it to be quite true at MS too. Yes, there were some people who seemed pretty boring and one-dimensional at MS, but there are those types of people everywhere. In my experiences at MS, there were many dynamic individuals who were very excited about what they worked on and had numerous hobbies and interests outside of computers.

    As I said earlier, though, I’ve never visited MS research (in fact, I hadn’t learned of their wealth of CS gurus until working towards my Masters Degree and reading papers from the likes of Hoare, Rashid, Lamport, etc.), so perhaps MS research is full of boring stiffs who are only interested in their 401k plans. But, to be honest, I doubt that this is the case.

  6. I work in a huge unix shop with hundreds and hundreds enterprise class machines. We do have a few windows servers that many people here refer to as “little toy boxes”. They run some non-mission critical apps and no one really cares if they are down. All this “innovation” by Microsoft scientists may impress an average CompUsa visitor but means nothing for the enterprise that requires high availability. It’s going to be a while before MS will be able to play even with the big boys.

  7. If they are non-mission critical, and no one really cares if they are down, why bother keeping them running?

  8. Er, in what sense is Microsoft — with, what is it, a 95% share of desktops? — not a big boy?

    Disclosure — Mac user.

  9. While doing time at those particular institutions (MIT 89-93, MSR 93-00) I crossed paths with really boring people at each, in just about equal proportions. I also met fascinating, intelligent, multi-disciplined friends in both places. In neither place was anyone on the “downslope of the IQ curve” unless you mean the right side of the bell.

    But you’re right: you won’t find many biologists or geologists at Microsoft, nor is Microsoft Research attempting to assemble a world-class collection of cocktail party guests. Perhaps when big minds choose to enter the “roach motel” they decide based on criteria similar to those listed by Scott and Pedro, rather than whether they can bring their dogs to work?

    (Interestingly, if you do a web search you’ll see that Paul Saffo has been repeating this particular witicism since at least early ’98. Maybe he should innovate a new catch phrase? Until he’s found something else to talk about, I sure wouldn’t want to get stuck next to him at a cocktail party…)

  10. Back to the dogs at the office argument: I got a yellow lab+beagle mutt that has a gas problem, if I take him to the office I will either be fired within the week or he will cause the dog-friendly policy to be withdrawn.

  11. 95% desktops – may be. But your data is processed and stored on the servers. With apps moving to the web a desktop is just a little more advanced than a dumb terminal. Why have chip on the shoulder, considering that half of those 95% is broken and crashes and the other half has to be babysitted and patched against endless security holes?

  12. Remind me again why should I bust my ass so my boss can make payments on his samurai village in Woodside/mega-yacht/Gulfstream V/da Vinci manuscript?

  13. ” All this “innovation” by Microsoft scientists may impress an average CompUsa visitor but means nothing for the enterprise that requires high availability” That is an odd quote. The average CompUSA visitor neither knows nor cares about the scientists just like Bell telephone customer could hardly be expected to have known or cared about Bell Labs (or substitute Xerox PARC or probably many others). But it became one of the great research institutions. I doubt that the builders of high capacity telephone switches knew or cared about Bell Labs just as you seem to have little regard for the scientists at Microsoft Research. I think the point Mr Saffo was making is that despite the wealth of talent Microsoft Research has not yielded the landmark developments like the transistor or later the discovery of the cosmic background radiation at Bell Labs. I don’t see how high availability enterprise computing would have anything to do with it.

Comments are closed.